Eoghan O'Neill (UCD)
will speak on
Autoregressive limited dependent variable models with Bayesian additive regression trees
Time: 3:00PM
Date: Thu 6th November 2025
Location: E0.32 (beside Pi restaurant)
[map]
Abstract: This paper describes a general approach to MCMC sampling for models with Bayesian Additive Regression Trees (ARBART) that split on the lag of a latent score. Filtering and smoothing methods are introduced for sampling of the latent scores. These results allow for estimation of autoregressive binary probit BART, ordered probit BART, and multinomial unordered probit BART models for time series and panel data. Further extensions to other models with nonlinear autoregressions of latent scores are also discussed.
(This talk is part of the Statistics and Actuarial Science series.)
PDF notice
Return to all seminars